Intelligent monitoring of drilling operations in sensitive environments (INDORSE)

Controlled releases of drill cuttings and mud into the sea are often a necessary component of drilling operations. It is important to minimise harm to sensitive species, such as cold-water corals and sponges, which might be exposed to the discharged material. Monitoring how this material is transported within the water, and eventually onto the seabed, is needed to continually assess the environmental impact from these types of drilling operations. Ocean forecasting and mathematical models play an important role in environmental monitoring, and real-time measurements are required to correct for model errors. Fixed sensor installations are expensive to deploy and maintain, and provide only fixed point or profile measurements of variables like ocean current or particle density and size distribution. In recent years, autonomous underwater vehicles (AUVs) have become a very useful alternative, providing opportunities for adaptive sampling, where units move to the locations where the most useful data can be found. The ambition of the INDORSE project has been to improve our knowledge about how to use AUVs to intelligently measure particle distributions in the water. The main goal of the project has been to develop knowledge and technology for near real-time and cost-efficient monitoring of particle dispersal during drilling operations. The subgoals have been to analyse the spatial and temporal dynamics of particle dispersal, to find a good balance between autonomy and communication during AUV operations, to demonstrate important parts of the system in the field, and to educate a PhD candidate and publish results in international journals. We have established systems for operationally running the ocean model SINMOD and the particle transport model DREAM. This means that daily simulations were conducted, using the most up-to-date data available from, e.g., the Norwegian Meteorological Institute. The models were set up to provide a forecast of approximately 48 hours, and this system was used in preparation for, and during, two field works in 2018 and 2020. The field works were conducted in Frænfjorden, where there is an active deposit site for mineral particles. The location was chosen due to its sheltered location, relatively short distance from Trondheim, and because the area and the deposit is known to us from earlier projects. The project has also done work on data assimilation of measurements into SINMOD, and development of particle transport algorithms. Model results from SINMOD and DREAM have been used to establish and train a so-called Gaussian model for the spatial distribution of particle concentrations in Frænfjorden. The Gaussian model was then used as a part of the autonomous control system on an AUV during the field works. We additionally conducted measurements of signal strength related to acoustic communication between the AUV and a modem. These data were used in further development of a model for predicting the signal strength based on hydrography (temperature and salinity profiles, which affect how sound is transmitted in the ocean). We have demonstrated that it is possible to set up an operational coupled model system with an ocean model and a particle transport model. When this model chain is forced with input data from the weather forecast, the system can give a prediction of particle concentrations 1-2 days ahead. We have also demonstrated the use of model data from ocean and particle transport models to train a Gaussian model, which in turn has been used in autonomous planning onboard an AUV. The project has so far financed (in part or in full) three journal papers, one book chapter, and one conference paper. Additionally, 3-4 further papers are in preparation, and are expected to be submitted in 2021 or early 2022. The PhD candidate started in August 2017, and is doing a 4-year PhD with 25% teaching, where the extra year is funded by NTNU. The PhD defence is planned for the first half of 2022. AUVs are fast becoming a cheaper and more accessible tool for marine monitoring. The project has developed knowledge about autonomous mission planning and measurement of particle concentrations by AUV. This has applications for monitoring of mineral particle concentrations, for example related to petroleum drilling or mining activity, but there are also applications within, e.g., marine biology. There are at present several groups at NTNU that work with autonomous mission planning for AUVs, with different applications in mind. Model development for particle transport, as well as measurements from the field works, help improve SINTEF's particle transport model DREAM. This model is actively used by the petroleum industry for modelling particle dispersal from drilling operations. Results from this project will also be used in future projects that look at environmental effects of mineral extraction (both land-based activities with deposits at sea, and sea-floor mining).

Arbeidet på ruteplanlegging er en viktig del av å gjøre AUVer til et anvendelig verktøy for miljøovervåkning. På sikt kan dette bli en standard prosedyre som kan stilles som krav i utslippstillatelser, for eksempel i mineralindustrien. Fordelen med en AUV er at den kan måle kontinuerlig (den må naturligvis lade batteriene, men f.eks. daglige måleserier er mulig), og den kan måle over større områder, og selv avgjøre hvor målingene bør tas for å redusere usikkerheten mest mulig. Slik sett er AUVen potensielt overlegen både manuelle kartlegginger, som typisk vil gjøres sjelden, og fastmonterte bøyer, som kun måler i et punkt eller en profil. Prosjektet har også bidratt til modellutvikling, samt gjort målinger i felt som vil bidra til fremtidig utvikling, innen modellering av spredning av mineralpartikler i havet. Dette gjør oss i stand til å gi mer realistiske prediksjoner av miljøeffektene av f.eks. petroleumsboring eller deponi av gruveavgang.

Petroleum well drilling activity on the Norwegian continental shelf is expected to remain high in the coming years, but there are concerns regarding the environmental impact of the discharge of drill cuttings and mud on sensitive species like cold-water coral reefs and sponges. For sensitive areas, the Norwegian Environment Agency recommends both dispersion modelling in the planning phase, and monitoring of sedimentation, turbidity and other variables in the execution phase. Effective monitoring will allow adjustments to be made during operations when required in order to reduce the risk of harm to the environment. The INDORSE project focuses on the spatial and temporal dynamics of particle dispersal from oil drilling, and methods and technology for observing the system. To adequately sample the local environment, this project calls for a swarm of autonomous mobile measurement platforms (AUVs) equipped with sensors measuring current, hydrography and suspended particle content utilizing adaptive sampling strategies. The interaction between the AUVs and the decision support system requires acoustic communication. INDORSE aims to achieve its objectives through 1) analysing the dynamics of the dispersal of drill cuttings and how to observe it, 2) utilizing state-of-the-art mathematical models of the process, 3) analysing the feasibility and quality of acoustic communication between AUVs and installations, and 4) developing, testing and demonstrating key elements of the monitoring system. The project will employ a PhD student who will make important contributions advancing our knowledge of the system dynamics, and establishing the requirements for the observation system. The impact of the project will be important steps towards high precision monitoring and forecasting of environmental risk of maritime operations in sensitive environments. This will lead to reduced environmental impact of offshore operations.